摘要
通过对上海黄金交易所现货黄金Au99.95品种2007年1月4日到2014年2月28日共1736个交易日的收盘价格的研究,发现上海黄金交易所黄金收益率序列具有长期记忆性,其残差具有明显的波动集聚性。利用时间序列的相关理论,并通过EVIEWS6.0及MATLAB软件对序列进行分析,建立ARFIMA-GARCH模型,此模型反映了黄金收益率序列的长记忆性和波动集聚性,并对黄金收益率序列进行预测。结果显示,预测结果与实际价格拟合度高,预测误差较小。该模型可以更加准确地描述黄金价格序列的动态特征,通过此模型可以使黄金投资者和生产者更加了解黄金价格序列的特点。
The gold return series has a long-term memory and its residuals generate significant volatility clusters have been founded. The ARFIMA-GARCH model of gold price is presented by time series theory based on the closing price of gold in stock named AU99.95 in Shanghai Gold Exchange during 2007-01- 04--2014-02-28, totally 1736 trading days. This model can fit the series precisely and can be showed by the empirical results, and the prediction error is small with software EVIEWS6. 0 and MATLAB. The model can be taken advantage to describe the production process of gold price dynamically. And it＇s beneficial to gold producers and investors to know more about the characteristics of the sequence of the gold price.
出处
《青岛大学学报（自然科学版）》
CAS
2014年第4期10-14,共5页
Journal of Qingdao University(Natural Science Edition)
作者简介
叶静，女，硕士研究生，主要研究方向：数理金融。
通讯作者：赵凯，男，教授，主要研究方向：小波分析，数理金融等。
